Not long ago, a network outage of one sort or another was an annoying but relatively minor problem at many businesses. If you lost your connection to the net, you just worked locally until the IT folks got it fixed.
This doesn't work anymore. The network fails and the point-of-sale terminal, formerly known as a cash register, is useless. The corporate data in a Web app on your screen is frozen. Your computer itself may suddenly become useless because it is really an image of a virtual machine running on a server somewhere in the cloud. Even a brief service outage can lead to serious losses of productivity, sales, and output. "Customers are more dependent on networks than ever before," says Christopher Kimm, vice-president for global network field operations at Verizon Communications. "We often see stark realization after a fault. Their exposure was greater than they realized."
This means the traditional model of network maintenance, waiting for something to break and then fixing it, is no longer adequate for many businesses, nor is the brute force approach of just building in more redundancy. Instead, a new breed of smart, proactive services is being employed to look after network health. In a report called "Smart" Services for Network Management Will Be Critical for Business Success in a Connected World," Forrester Research says: "Through continual monitoring of the network and devices connected to it, these technologies can identify stresses and potential points of failure and predict where problems could surface. In short, they allow firms to understand the usage, demands, and potential weak spots, which can be used to anticipate and prepare for future strains as new services and activities are added to the network."
Continuous monitoring of networks, generally relying on extensive automation through software, is not a task for which most enterprises are particularly well equipped. So it's not surprising that growing demand for these full-time network health checks is being met by third parties, including Cisco, offering network smart services.
Verizon notes that smart services work best when a high degree of reliability and resilience is built into the network. "We put a lot of energy into the design portion," says Kimm. "With enterprise customers, this is a relatively easy conversation to have. We think first and foremost that the network, in the broadest possible sense, has to be designed with a resilience level appropriate to the risk of the business, the economic opportunity, and what the customer can afford."
Another key element of the smart services offering is a service level agreement (SLA) that creates the right incentives. "With customers with whom we get very engaged, we have to deliver very tailored SLAs," says Kimm. "We can offer very high SLA levels because we can deliver. We have confidence in the tools and their ability to help us."
Cisco's smart services offering, delivered direct or through partners such as IBM Global Services, has similar goals and, just as Verizon's support encompasses much more than the wide-area networking links and internet backbone it supplies, Cisco's management goes beyond its routers, switches, and other gear. "Most of the internet is stuff we know about," says Nick Earle, senior vice president for worldwide services sales .
Cisco's goal is to use smart software to reduce network management costs while improving performance and reliability. "Our customers have to increase the value of their networks, but they are looking for a 20% or more reduction in cost. Most services are based on people and the cost of people goes up every year. We need a different approach – one that leverages automated software capabilities. "
In addition to monitoring, Earle sees getting the network right in the first place as a key to success. Cisco smart services are based on the Information Technology Infrastructure Library, the industry gold standard for configuration management.
Like Kimm, Earle sees the increasing software intelligence built into networks as the facilitator of more efficient and reliable management, especially by identifying the small number of problems that produce the majority of faults. "Software can sense all the patterns of configuration in your network. We can go one step further and show a correlation between configuration patterns and network incidents. It's always a Pareto*. Individuals can't do that."
In the end, says Verizon's Kimm, networking is becoming a vital utility, but one that people who are not primarily in the networking business don't want to much think about: "People think of it as being like electricity and they think they don't have to know how it works. We take it for granted."
*-The Pareto principle, named for the Italian economist Vilfredo Pareto, holds that for many activities, a small proportion of causes is responsible for a large proportion of outcomes. In engineering, Pareto analysis seeks to find and correct those few conditions that are responsible for a large percentage of faults.
The contents or opinions in this feature are independent and do not necessarily represent the views of Cisco. They are offered in an effort to encourage continuing conversations on a broad range of innovative technology subjects. We welcome your comments and engagement.
We welcome the re-use, republication, and distribution of "The Network" content. Please credit us with the following information: Used with the permission of http://thenetwork.cisco.com/.